JP2000101860A - Image correction method, image correction device and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To conduct color correction and density correction without causing any deterioration in the image quality such as change in principal gradation, spotting of white portions in a highlight part and blackish white-background in the image. SOLUTION: After calculating a correction value (gain) for color correction and density correction, as shown in Figure (B), a transformation curve on the outside of a prescribed range (the range of about 0.5 to 1.0 in antilogarithm or of about -0.3 to 0 in logarithm in Figure (B)) is set with a gradient of unity on a log-log chart where the X axis is the logarithm of an input value and the Y axis is the logarithm of an output value, the set transformation curve is moved in parallel depending on the gain so that the gradient is unchanged, and the transformation curve within the prescribed range is decided so that a value corresponding to a brightest point in an input value is converted into a value corresponding to a brightest point in an output value, a middle part of the curve is changed smoothly and the curve connects to the transformation curve at the outside of the prescribed range. Then object reflectance data obtained from image data are converted according to image data conversion conditions specified by the above transformation curve.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image correction method, an image correction apparatus, and a recording medium, and more particularly, to an image correction method for correcting at least one of the color and density of an image, and an image correction to which the image correction method can be applied. The present invention relates to an apparatus and a recording medium on which a program for causing a computer to execute the image correction method is recorded.

[0002]

2. Description of the Related Art It is necessary to perform color correction such as white balance adjustment and density correction for optimizing the density of the entire image on image data obtained by imaging with a video camera or imaging with a digital still camera. Generally, color correction and density correction of an image are performed on a plurality of channels (for example, R, G,
B) The image data is sampled for each channel, an average value is calculated for each channel based on the data of each channel obtained by the sampling, and based on the calculated average value, for example, for color correction, the average for chromaticity is calculated. For the density correction, the average value for the luminance is set to an intermediate value (for the reflection density, “0.7
5), and a value such as “118” is used for image data representing the density of each pixel in 8 bits. Has been converted.

The conversion of image data corresponding to color correction and density correction is often performed according to the following conversion formula.

[0004]

(Equation 1) In the above equations (1) and (2), (R, G, B)
Image data before correction (R of each pixel of the image, representing the value of G, B), (R ' , G', B ') image data after _{correction, (k r, k g,} k b) Represents a conversion coefficient of each channel (corresponding to correction values of color correction and density correction). (1)
The expression is a conversion expression for adjusting the gain, and the expression (2) is a conversion expression for adjusting the offset. Either one of the expressions (1) and (2) is used or a combination of the two expressions is used (referred to as affine transformation). Thus, conversion of image data corresponding to color correction and density correction can be performed.

[0005]

However, when the image data to be processed is data with a small number of bits per pixel / channel (for example, 8 bits), that is, image data with a small (narrow) dynamic range. When the color correction and the density correction are performed according to the formulas (1) and (2), the image represented by the corrected image data should be reproduced as, for example, the highlight portion becomes white or the image is originally white. If the area (white area) is blackish,
There is a problem that the image quality becomes unfavorable.

In order to solve this problem, the maximum value ("255" in 8-bit image data) in the image data is converted into the maximum value, and the minimum value ("0") in the image data is converted into the minimum value. It has been proposed to convert image data according to a conversion condition that defines a conversion characteristic in an intermediate value region so as to satisfy the conditions and perform desired color correction and density correction. Another problem arises in that the gradation in the density range (mainly the intermediate density range) (hereinafter referred to as the main gradation) changes.

The present invention has been made in consideration of the above facts, and has image quality such as a change in main gradation of an image, a highlight portion in an image being white, and a white background in an image being blackish. It is an object of the present invention to provide an image correction method, an image correction apparatus, and a recording medium that can perform color correction and density correction without lowering.

[0008]

In order to achieve the above object, an image correction method according to the first aspect of the present invention is based on a correction amount for correcting at least one of a color and a density of an image to be corrected. The value corresponding to the brightest point in the input signal is converted to the value corresponding to the brightest point in the output signal, and the slope of the change in the logarithmic value of the output signal value with respect to the change in the logarithmic value of the input signal value. Is set to be approximately 1 except for a predetermined area near the signal value corresponding to the brightest point, and the relationship with the light intensity of each pixel of the correction target image represents a linear physical value. The image data is converted according to the set conversion conditions.

According to the first aspect of the present invention, a value corresponding to the brightest point in the input signal is determined based on the correction amount for correcting at least one of the color and the density of the image to be corrected. In addition to being converted to the value corresponding to the brightest point, the slope of the change in the logarithmic value of the output signal value with respect to the change in the logarithmic value of the input signal value is substantially the same except for a predetermined region near the signal value corresponding to the brightest point. The conversion condition is set to be 1. Specifically, the above conversion condition is, for example, a two-dimensional bilogarithmic chart (bilogarithmic plot) in which the logarithmic value of the input signal value is taken on one coordinate axis and the logarithmic value of the output signal value is taken on the other coordinate axis. Then, a conversion curve (straight line) in an area outside the predetermined area is set so that the inclination is substantially 1, and the conversion curve is translated by a movement amount corresponding to the correction amount so that the inclination does not change. Is smoothly connected to the conversion curve outside the predetermined area, and the value corresponding to the brightest point of the input signal (for example, 0 or 255 when the input signal is 8-bit data) is the output signal. (For example, if the output signal is 8-bit data, 0 or 255) can be set. Note that the correction amount for correcting at least one of the color and the density of the correction target image can be obtained, for example, by calculation from the image data.

Generally, since the human sensory intensity is proportional to the logarithm of the physical intensity of the stimulus, as described above, the slope of the change in the logarithmic value of the output signal value with respect to the change in the logarithmic value of the input signal value is within a predetermined range. By setting the conversion condition so as to be approximately 1 except for the above, even if the conversion of the image data according to the conversion condition (that is, at least one of the color and the density of the correction target image) is performed, the converted image It is possible to prevent the main gradation of the image represented by the data from being recognized as having changed. Further, since the conversion condition is set so that the value corresponding to the brightest point of the input signal is converted to the value corresponding to the brightest point of the output signal, the conversion condition in the image represented by the image data after conversion is set. It is also possible to avoid that the highlight portion becomes white and the white background in the image becomes blackish.

According to the first aspect of the present invention, the relationship between the light intensity of each pixel of the image to be corrected and the light intensity is a linear physical quantity (the light intensity itself may be the light intensity, the light reflectance may be the same).
Is converted according to the conversion conditions set as described above, so that the main gradation of the image changes, the highlight portion in the image becomes white, and the white background in the image becomes black. Color correction and density correction can be performed without causing image quality deterioration such as tinge.

When image data representing a value of a physical quantity having a non-linear relationship with light intensity is input as a physical quantity for each pixel of an image to be corrected, the image data is also used as an image data according to the present invention. (Image data representing a linear physical value in relation to the light intensity of each pixel of the correction target image), and converting the converted image data in accordance with the conversion conditions according to the present invention. The color correction and the density correction can be performed without lowering the image quality, such as that the highlight portion becomes white or the white background in the image becomes blackish.

By the way, the inventor of the present application has made the predetermined area (the area where the slope of the change of the logarithmic value of the output signal value with respect to the change of the logarithmic value of the input signal value is not substantially 1) in the conversion condition according to the present invention various sizes. , A plurality of types of conversion conditions were set, image data was converted using each conversion condition, and an experiment was performed to evaluate the image quality of the image represented by the converted image data. As a result, when the input signal value is normalized within the numerical range of 0 to 1, the logarithmic value of the input signal value is within approximately 0.5 (however, a common logarithmic value) from the signal value corresponding to the brightest point. It has been found that if the area corresponding to the range is set as the predetermined area, visually favorable image quality is obtained.

Therefore, according to a second aspect of the present invention, in the first aspect of the present invention, the predetermined area is set such that the input signal value is from 0 to 1
Are standardized within the range of from the signal value corresponding to the brightest point to the logarithmic value of the input signal value within a range of approximately 0.5 or less. By defining the range of the predetermined area as described above, the image quality of the image represented by the converted image data can be visually enhanced.

According to a third aspect of the present invention, a value corresponding to the brightest point in the input signal is determined based on a correction amount for correcting at least one of the color and the density of the image to be corrected. The output signal is converted to a value corresponding to the brightest point, and the slope of the change in the logarithmic value of the output signal value with respect to the change in the logarithmic value of the input signal value is a predetermined value near the signal value corresponding to the brightest point. Approx. 1 excluding area
Setting means for setting a conversion condition so that the image data representing a value of a physical quantity having a linear relationship with the light intensity of each pixel of the image to be corrected is converted in accordance with the conversion condition set by the setting means. And conversion means.

According to the third aspect of the present invention, the value corresponding to the brightest point in the input signal is output by the setting means based on the correction amount for correcting at least one of the color and the density of the image to be corrected. The signal is converted into a value corresponding to the brightest point of the signal, and the slope of the change in the logarithmic value of the output signal value with respect to the change in the logarithmic value of the input signal value is a predetermined area near the signal value corresponding to the brightest point. Approximately 1 except
The conversion condition is set so that the image data representing the value of the physical quantity having a linear relationship with the light intensity of each pixel of the correction target image is converted by the conversion unit according to the conversion condition set by the setting unit. Therefore, in the same manner as in the first aspect of the present invention, there is no change in image quality such as a change in the main gradation of the image, a highlight portion in the image becoming white, and a white background in the image becoming blackish. Correction and density correction can be performed.

According to a fourth aspect of the present invention, a value corresponding to the brightest point in the input signal is output based on the correction amount for correcting at least one of the color and the density of the image to be corrected. The signal is converted into a value corresponding to the brightest point of the signal, and the slope of the change in the logarithmic value of the output signal value with respect to the change in the logarithmic value of the input signal value is a predetermined area near the signal value corresponding to the brightest point. A first step of setting a conversion condition so as to be approximately 1 except for the step of converting image data representing a value of a physical quantity having a linear relationship with the light intensity of each pixel of the correction target image according to the set conversion condition. A program for causing a computer to execute a process including the second step of conversion is recorded.

The recording medium according to the fourth aspect of the present invention includes:
A process including the first step and the second step,
That is, since a program for causing a computer to execute the processing according to the image correction method according to the first aspect of the present invention is recorded, the computer reads out and executes the program recorded on the recording medium. In the same manner as in the first aspect of the present invention, color correction and density correction can be performed without causing image quality deterioration such as a change in the main gradation of the image, a highlight portion in the image becoming white, and a white background in the image becoming blackish. It can be carried out.

[0019]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the present invention will be described below in detail with reference to the drawings. FIG. 1 shows an image processing system 10 according to the present embodiment. The image processing system 10 is a film image recorded on a photographic material (hereinafter simply referred to as a photographic film) such as a photographic film (for example, a negative film or a reversal film). Digital lab system 1 configured to be able to process at a high speed reading and recording on a photographic paper
2, the image data exchange 14 and the interface (I /
F) The input device group 18 and the output device group 2 via the circuit 16
0 is connected.

The input device group 18 includes the image data exchange 14
, And various input devices for inputting image data. The input devices constituting the input device group 18 include, for example, a magnetic disk such as a floppy disk (FD) and a CD.
-R and other optical discs, magneto-optical discs (MO), digital still cameras (DSC: hereinafter simply referred to as "digital cameras"), and PC cards and IC cards (hereinafter collectively referred to as "digital camera cards") ) Is set, and the information storage medium reading device 22 (see FIG. 2) for reading and inputting image data stored in the set information storage medium.
Alternatively, a communication control device (not shown) that receives and inputs image data transmitted from another information processing device connected via a communication line can be applied.

The output device group 20 includes various output devices that perform image output processing based on the output image data transferred from the image data exchange 14. An output device constituting the output device group 20 is, for example, an information storage medium writing device (for example, a CD as an information storage medium) that writes image data to an information storage medium (for example, a CD-R) as an image output process. 2 shows a CD-R writer 24 for writing image data to R), an image display device for displaying an image on a display means such as a display as an image output process, and a communication line as an image output process. A communication control device or the like for transmitting image data to another information processing device connected via the communication device can be applied.

By the way, the file structure of the image data input from each of the input devices constituting the input device group 18 is not constant and is often different from each other. Therefore I
When image data is input from the input device, the / F circuit 16 determines the file structure of the input image data, converts the input image data into a predetermined file structure, and inputs the file structure to the image data exchange 14. The output device pre-defines the file structure of image data transferred from the outside, but this file structure is often different for each output device constituting the output device group 20. Therefore, when the image data is transferred from the image data exchange 14 to the output device, the I / F circuit 16 converts the file structure of the transferred image data into a file structure corresponding to the destination output device.

The digital lab system 12 includes the scanner 3
0, an image processing device 32 and a printer 34 are connected in series. The scanner 30 has a reading sensor such as an area CCD sensor, and reads a film image recorded on a photographic film by the reading sensor. The image data obtained by reading the film image is output to the image processing device 32 and used for image output processing (recording of an image on photographic paper) by the printer 34, but is used for image output processing by an output device other than the printer 34. The image data instructed to be used is also output to the image exchange 14.

The image processing device 32 performs pixel density conversion, color conversion, and super-low-frequency luminance component conversion on the input image data as image processing for exposing and recording an image on photographic paper with appropriate image quality. Hyper tone processing to compress gradation,
Hyper-sharpness processing for enhancing sharpness while suppressing graininess, special image processing (for example, correction of image quality deterioration or red-eye correction due to LF lens aberration for film images captured and recorded by LF (film with lens)) And various image processing circuits (not shown) for performing various image processing. The image processing device 32 calculates processing conditions for image processing performed by each image processing circuit. Each image processing circuit performs various types of image processing on the image data according to the calculated processing conditions, and the image data after the image processing is output to the printer 34 as recording image data.

The printer 34 includes R, G, and B laser light sources and a laser driver (not shown) for controlling the operation of the laser light sources (not shown).
The G and B laser beams are modulated by the input image data for recording, and the modulated laser beams are scanned on photographic paper. Thus, the image is exposed and recorded on the photographic paper. The photographic paper on which the image has been exposed and recorded is sent to a processor unit (not shown) and subjected to color development, bleach-fixing, washing and drying, and the image recorded and exposed on the photographic paper is visualized. The image data transferred from the image data exchange 14 to the printer 34 is also used for modulating the laser beam, that is, for exposing and recording the image on photographic paper in the same manner as described above.

As shown in FIG. 1, the image data exchange 14 has a CPU 78, a ROM 42, a RAM 44, and input / output ports 46A and 46B connected to each other via a bus 48. It includes a processing device, a hard disk device 50 containing a large-capacity information storage medium (hard disk) and connected to a bus 48, and a CD-ROM driver 54 for reading programs and the like from a loaded CD-ROM 52. The scanner 30 and the printer 34 of the digital lab system 12 are connected to the input / output port 46A, and the input device group 18 and the output device group 20 are connected to the input / output port 46B via the I / F circuit 16. .

The image data exchange 14 temporarily stores the image data input from the scanner 30 and the input devices of the input device group 18 in the internal hard disk of the hard disk device 50. Therefore, the built-in hard disk of the hard disk device 50 functions as a spool 60 (see FIG. 2) for storing image data input to the image data exchanger 14. Further, the image data exchanger 14 temporarily stores the input image data in the spool 60 after adding property information indicating an attribute or the like of the image data before temporarily storing the image data in the spool 60.

The built-in hard disk of the hard disk device 50 stores various image processing programs for performing various types of image processing on image data. These programs are selectively executed at a predetermined timing (at least one of timing before temporarily storing the image data in the spool 60 and after reading the image data temporarily stored in the spool 60), and Performs various image processing. As described above, the image data exchange 14 is an image processing engine 62 that performs various image processing.
(See FIG. 2).

In the present embodiment, as the image processing for the image data, as shown in FIG.
"Pixel density conversion" for converting image data into different color data, "color space conversion" for converting to image data in a different color space, "data compression (or decompression)", a predetermined format called FlashPix (a plurality of different resolutions different from each other) (Fl) that performs conversion (or inverse conversion) into image data that includes (pixel density) image data and that is obtained by dividing image data of each resolution into a plurality of small areas (called tiles).
ashPix Formatting ”,“ D ”which is image quality improvement processing for image data obtained by imaging with a digital camera.
"SC set up", "sharpness correction" to improve image sharpness, "digital watermark" for embedding predetermined digital watermark data to prevent unauthorized duplication of image data, etc. Various image processes such as "Composite" for generating image data of one image (for example, image data for creating a New Year's card, etc.) are prepared.

Among the above various image processing, "DSC set u
“p” includes a color correction / density correction process (to be described in detail later) to which the image correction method according to the present invention is applied, and a color conversion process for increasing the color of an image. CPU 40 of image data exchange 14 performs color correction and density correction processing
The color correction / density correction program to be executed by the CPU 40 is initially stored in the CD-ROM 52 together with the program for causing the CPU 40 to execute other image processing. When the CD-ROM 52 is loaded into the CD-ROM driver 54 and transfer (install) of a program from the CD-ROM 52 to the image data exchange 14 is instructed,
A color correction / density correction program and other programs are read from the CD-ROM 52 by the CD-ROM driver 54 and stored in the internal hard disk of the hard disk device 50.

Then, when it is time to execute the color correction / density correction processing, the color correction / density correction program is read from the internal hard disk of the hard disk device 50 and stored in the RAM 44 (when the power of the image data exchange 14 is turned on). At this time, the program for each image processing may be read and stored in the RAM 44).
Performed by U40. Thereby, the image data exchange 14 functions as an image correction device according to the present invention.
Note that other image processing programs are read and executed in the same manner as described above.

As described above, the CD-ROM storing the color correction / density correction program and other image processing programs is stored.
The ROM 52 and the internal hard disk of the hard disk device 50 correspond to the recording medium of the present invention.

Next, as an operation of the present embodiment, the information storage medium reading device 22 (and the CD-R
The transfer of image data from the information storage medium reading device 22 to the printer 34 of the digital lab system 12 will be described with an example in which the writing device 24) is connected (see FIG. 2).

In the image processing system 10, an FD or MO storing image data processed by an information processing device such as a personal computer owned by the user is brought in and requested to make a print, or an image is taken by a digital camera. The digital camera card storing the image data obtained by the above is brought in and a print creation request is made, or the image data exchange 14,
There is a case where a CD-R on which image data transferred via the CD-R writing device 24 is written is brought in and a request for print creation (printing) is made.

In this case, the information storage medium brought in by the user is set in the corresponding information storage medium reading device 22 (any one of FD drive, CD drive, MO drive, card reader, etc.), and the information storage medium is set. The information storage medium reading device 22 reads out the image data to be processed from the set information storage medium, and then converts the read image data into attribute information representing various attributes of the image data to be processed, and the attribute information of the image data. Output destination is printer 34
Is transferred to the image data exchange 14 together with the information indicating that

The image data transferred from the information storage medium reading device 22 is input to the image data exchange 14 after being converted into a predetermined file structure by the I / F circuit 16. The image processing engine 62 of the image data exchange 14 is configured such that the input source of the image data is the information storage medium reading device 22 and the information indicating the output destination of the image data input together with the image data is used to input the image data to be processed. Is the printer 34
When the image data to be output to the spool 60 is recognized, the attribute of the image data to be processed depends on the input source (the type of the information storage medium reading device 22) before the image data to be processed is stored in the spool 60. And optimal image processing according to the output destination (printer 34).

Here, the input source information storage medium reading device 2
If the device 2 is a device (card reader) that reads out image data from a digital camera card, the image processing engine 92 generates the input image data by imaging with a digital camera and stores the image data in the digital camera card ( In this case, it is determined that the image data is compressed (stored after being compressed), and the compressed image data is decompressed and converted into image data having a resolution (pixel density) suitable for recording the image on photographic paper. And image processing such as "DSC set up" including color correction / density correction processing and "sharpness correction" for improving image sharpness.

Hereinafter, the image data to be processed is obtained by encoding the image data read from the digital camera card (the values of the R, G, and B component colors (each channel) for each pixel according to predetermined conversion characteristics). Image data (R, G, B))
, The color correction / density correction processing executed by the image processing engine 62 (the CPU 40 of the image data exchange 14) executing the color correction / density correction program will be described with reference to the flowchart of FIG. .

That is, in step 100, the image data (R, G, B) is sampled, and in the next step 102
Then, the image data (R,
G, B), a predetermined image characteristic amount for color correction is calculated, a predetermined image characteristic amount for density correction is calculated, and color correction and density correction are performed based on the calculated image characteristic amounts. Correction value (gain for image data) is calculated for each channel. It is preferable that the sampling for color correction and the calculation of the image feature quantity and the sampling for density correction and the calculation of the image feature quantity be performed separately as follows.

For example, for the sampling for density correction and the calculation of the image feature amount, the image data (R, G,
B) sampling the data of all the pixels from
For each ₀ (for example, about 100) pixels, an averaged sampling value is obtained for each channel. (A sampling value (digital code value) of M ₀ pixels is converted into a subject reflectance to calculate an average value. (Inverse conversion to a digital code value) is repeated, and an average luminance value [v] is calculated from the averaged sampling value for each channel obtained for each of the M ₀ pixels as an image feature amount for density correction.

Further, for example, for sampling for color correction and calculation of image feature amount, image data (R,
G, S ₀ or from B) (e.g. the data of each channel of one pixel for each pixel of the approximately 100) repeatedly sampling (selection), as the image feature amount for color correction, for each channel The average values [r], [g], and [b] of the image data sampling values (digital code values) are calculated. Then, based on the average value [r] [g] [b] of the image data sampling value for each channel and the above-described average luminance value [v], correction for each channel for color correction and density correction is performed. Value (gain α _R for image data,
α _G , α _B ) are calculated.

In the above-described calculation of the averaged sampling value, M used for calculation of one averaged sampling value is used.
_Since the area of the region on the original image corresponding to ₀ pixels is large, the change in the luminance at the point where the high luminance point or the low luminance point exists in the original image is averaged, and the value representing the intermediate density becomes It is set as an averaged sampling value. By calculating a correction value using the average luminance value [v] obtained from such an averaged sampling value, it is possible to obtain a correction value that enables highly accurate density correction.

In the data obtained by sampling for color correction, since the area of the region on the original image corresponding to the sampled pixel is very small, a high brightness point or a low area where the area in the original image is very small is small. It contains a high percentage of data on luminance points and other locations that should be neutral colors in the original image. Therefore, by calculating a correction value using such an average value [r] [g] [b] of each sampling value for each channel, it is possible to obtain a correction value capable of performing highly accurate color correction.

In the next step 104, step 102
The correction value (gain α _R ,
Based on α _G , α _B ), color is applied to image data to be corrected (specifically, object reflectance data (r, g, b) representing an object reflectance value having a value within a range of 0 to 1). Image data conversion conditions for performing correction and density correction are set for each channel. Specifically, the image data conversion condition can be set by defining a conversion curve as described below, for example.

That is, as shown in FIG. 4B, the logarithm of the input image data value (input signal value) before conversion is plotted on the horizontal axis, and the output image data value (output signal value) after conversion is plotted on the vertical axis. 2), a two-dimensional logarithmic chart taking the logarithmic value of the subject value (note that in FIG. 4B, the subject reflectance value (exact numerical value) represented by the subject reflectance data to be converted is given along each coordinate axis. 4), a region outside a predetermined region (in FIG. 4B, a range of approximately 0.5 to 1 in a subject reflectance value (true numerical value) (a range of approximately −0.3 to 0 in a logarithmic value of the subject reflectance value)). That is, FIG.
In (B), a conversion curve (straight line) in a range of approximately 0.5 or less in object reflectance value (exact numerical value) (in a range of approximately -0.3 or less in logarithmic value of the object reflectance value) is set so that the slope becomes 1. (For example, a straight line of input image data value = output image data value (a straight line of gain = 1.0 shown in FIG. 4B is set).

Next, the set conversion curve is corrected with correction values (gains α _R , α
_G , α _B ) in parallel along the direction perpendicular to the horizontal axis. The movement of the conversion curve is, specifically, the gain α
Can be moved so as to approach the horizontal axis as the value of becomes smaller, and move away from the horizontal axis as the value of the gain α becomes larger (in FIG. 4B, the gain becomes “1.3” or “1.0”).
Conversion curves at each value of "0.7" are shown.)

After the conversion curve outside the predetermined area is determined as described above, the conversion curve inside the predetermined area is then converted to the value corresponding to the brightest point in the input image data (the true value in FIG. 4B). Is converted to a value corresponding to the brightest point in the output image data (in FIG. 4B, similarly, 1.0 is an exact value and 0 is a logarithmic value), and the middle part is smooth. To be smoothly connected to the conversion curve outside the predetermined area.

As a result, as shown in FIG. 4B, for example, the value corresponding to the brightest point in the input image data values is converted into the value corresponding to the brightest point in the output image data. In addition, the slope of the change in the logarithmic value of the output image data value with respect to the change in the logarithmic value of the input image data value is set to 1 except for a predetermined area near the image data value corresponding to the brightest point. Can be obtained in accordance with the value of the gain α. In FIG. 4A, the three conversion curves shown in FIG. 4B are plotted with the abscissa representing the true value of the input image data value before conversion, and the ordinate representing the output image data after conversion. For reference, a conversion curve when plotted on a two-dimensional antilogarithmic chart in which the exact numerical values are taken is shown.

Then, the above-mentioned conversion curve is set for each channel. For example, conversion (color correction and density correction) for the image data to be corrected is performed using a look-up table (L).
UT), image data conversion conditions can be set by defining conversion data to be set in the LUT based on a conversion curve for each channel. For example, conversion to image data to be corrected is performed by a functional expression , Image data conversion conditions can be set by generating a function equation that approximates a conversion curve for each channel.

Instead of setting the image data conversion conditions as described above, the image data conversion conditions when the gain α (correction value) is each value are determined and stored in advance, and at step 102 The image data conversion condition used for the conversion of the image data may be selectively set according to the obtained gain α (correction value). Further, the algorithm for setting the image data conversion conditions set above is merely an example, and it goes without saying that any algorithm that can obtain the same conversion conditions can be adopted.

By the way, image data stored in a digital camera card by photographing with a digital camera is as follows.
In general, according to the recommendation of the International Radio Communication Advisory Committee 709, the object reflectance X (takes a value of 0 to 1) is calculated using an 8-bit digital code value (0 to 0) by a function F (X) shown in the following equation (3). 255 (takes a value of 255) (coded as an integer).

[0052]

(Equation 2) As is apparent from equation (3), the relationship between the digital code value of each pixel of the image data from the digital camera and the light intensity is non-linear, and the color correction and density are directly applied to the image data from the digital camera. If the correction is performed, the gradation of the image changes with the correction.

For this reason, in the next step 106, the image data (R, G, B) is
Subject reflectance data (r, g, b) representing subject reflectance
Convert to This conversion can be realized by performing an inverse conversion of the function F (conversion by the function F ⁻¹ ) (see the following equation).

R ← F ⁻¹ (R), g ← F ⁻¹ (G), b ← F ⁻¹ (B) Thus, the value of each pixel of the image data (R, G, B) for each channel The (digital code value) is converted into a subject reflectance value having a linear relationship with the light intensity. The subject reflectance data (r, g, b) corresponds to “image data representing a value of a physical quantity whose relationship with the light intensity of each pixel of the correction target image is linear” according to the present invention.

In step 108, the previous step 104
According to the image data conversion conditions set for each channel, the subject reflectance data (r, g,
b) is converted. As a result, the object reflectance data (r, g, b) is set in step 1 for each channel.
Color correction and density correction corresponding to the correction values (gains α _R , α _G , α _B ) calculated in 02 are performed.

As described above, the image data conversion conditions are set so that the slope of the change in the logarithmic value of the output image data value with respect to the change in the logarithmic value of the input image data value is 1 except for a predetermined area. Therefore, even if the image represented by the converted image data is observed, for example, by recording the image on a recording material, it is clear from the fact that the human sensory intensity is proportional to the logarithm of the physical intensity of the stimulus. In this way, it is not visually recognized that the main gradation of the image represented by the converted image data has changed.

The image data conversion condition is set so that the value corresponding to the brightest point in the input image data is converted into the value corresponding to the brightest point in the output image data. The highlighted portion in the image represented by the converted image data does not appear white or the white background in the image does not become black. Further, the image data conversion condition according to the present embodiment is that the subject reflectance value (exact value) is approximately 0.5 to
An area corresponding to a range of 1 (a range of approximately -0.3 to 0 as a logarithmic value of the object reflectance value) is a predetermined area, and is included in the numerical range of the predetermined area according to claim 2. For this reason, as apparent from the experimental results described later, the image represented by the converted image data can be made a visually preferable image.

In step 110, the object reflectance data (r ', g', b ') obtained by the above-described color correction and density correction are converted into image data (R', G ',
B ′). This conversion can be performed by using the function F shown as the equation (3) (see the following equation).

R ′ ← F (r ′), G ′ ← F (g ′),
B ′ ← F (b ′) Accordingly, the main gradation of the image represented by the image data does not change, the highlight portion in the image does not appear white, and the white background in the image does not become blackish. , Image data (R ′, G ′, B ′) on which color correction and density correction have been performed. Various conversions for image data in the above-described color correction / density correction processing are summarized as follows.

[0060]

(Equation 3) In the above, the case where the color correction and the density correction are performed on the subject reflectance data (r, g, b) has been described. However, the present invention is not limited to this, and the subject reflectance data (r, g, b) is not limited thereto.
g, b) is converted to tristimulus value data (X, Y, Z) representing tristimulus values in the XYZ color system, and the tristimulus value data (X, Y, Z) is converted in the same manner as in step 110 above. After performing color correction and density correction, the tristimulus value data (X ′, Y ′, Z ′) after color correction and density correction are converted into object reflectance data (r ′, g ′, b ′). , Image data (R ′, G ′, B ′). From the object reflectance data (r, g, b), tristimulus value data (X,
(Y, Z) can be realized by a matrix operation as shown in equation (4). The object reflectance data (r ′) is obtained from the tristimulus value data (X ′, Y ′, Z ′) that has undergone color correction and density correction. ', G', b ') can be realized by a matrix operation as shown in the following equation (5).

[0061]

(Equation 4) Various conversions for the image data in this case are summarized as follows.

[0062]

(Equation 5) If the image data to be processed is image data (Y, Cr, Cb) in which the values of each channel of luminance Y and color difference Cr, Cb are coded according to a predetermined conversion characteristic for each pixel, From image data (Y, Cr, Cb) to image data (R,
G, B), color correction and density correction can be performed in the same manner as the color correction / density correction processing described above. In this conversion, first, luminance data Y and color difference data Cr, C
Data Luma, Chroma1, Chroma2 are obtained from b.

Y = (255 / 1.402) Luma Cr = 111.40 Chroma1 + 156 Cb = 135.64 Chroma2 + 137 (6) Then, according to the following equation (7), the data Lum is calculated for each pixel.
Data R, G, and B are obtained from a, Chroma1, and Chroma2. Thus, image data (R, G, B) can be obtained. This transformation is a kind of affine transformation.

[0064]

(Equation 6) Conversion of the corrected image data (R ′, G ′, B ′) into image data (Y ′, Cr ′, Cb ′) is performed by first using the data R ', G', B '
Data Luma, Chroma1, Chroma2 from

[0065]

(Equation 7) Then, from the data Luma, Chroma1, Chroma2, the previous (6)
According to the formula, the luminance data Y ′ and the chrominance data C
Find r ', Cb'. Thereby, the color correction and the density correction are performed without changing the main gradation of the image represented by the image data, the highlight portion in the image becoming white, and the white background in the image becoming blackish. The obtained image data (Y ′, Cr ′, Cb ′) can be obtained. Various conversions for image data when image data (Y, Cr, Cb) are input are summarized as follows.

[0066]

(Equation 8) In the above description, the conversion from the image data (R, G, B) to the subject reflectance data (r, g, b) and the inverse conversion are performed based on the expression (3). However, the present invention is not limited to this. Instead, for example, it may be performed based on the following equation (9).

[0067]

(Equation 9) Further, in the above, the recording medium according to the present invention is a CD-R
The OM 52 and the built-in hard disk of the hard disk device 50 have been described as examples, but the present invention is not limited to this. For example, a magnetic disk such as a floppy disk, an optical disk such as a CD-R, a magneto-optical disk such as an MO, a memory It goes without saying that various information storage media such as cards and IC cards can be applied as the recording media according to the present invention.

In the above description, C of the image data exchange 14
The PU 40 performs the color correction / density correction processing by executing the color correction / density correction program. However, the present invention is not limited to this. Processing circuit), and the image processing circuit may perform color correction and density correction processing.

In the above description, the case where the color correction value and the density correction value are automatically calculated has been described. However, the present invention is not limited to this. For example, when an image is displayed on display means such as a display. Alternatively, color correction and density correction may be performed based on the color correction value and the density correction value determined by the operator testing the display image.

In the above description, the image data generated by the image pickup by the digital camera is processed. However, the present invention is not limited to this. For example, the image data obtained by reading a film image recorded on a photographic film can be obtained. The present invention is applicable to color correction and density correction for image data and the like. Further, in the above description, the color correction and the density correction are respectively performed, but it goes without saying that only one of them may be performed.

[0071]

Next, the results of experiments conducted by the present inventors will be described. The inventor of the present application performed the following experiment in order to confirm an appropriate range of the predetermined region (an appropriate boundary between the predetermined region and the outside of the predetermined region) according to the present invention.

That is, a predetermined subject is photographed by a digital camera (Finepix 700 (trade name) manufactured by Fuji Film), and image data (R, G, B) obtained by the photographing is taken.
Is converted into subject reflectance data (r, g, b) by the inverse conversion of the function F (X) described above, and then is converted according to the conversion function f (x) to perform density correction. The object reflectance data (r ′, g ′, b ′) is converted to a function F
The image data is converted into image data (R ', G', B ') by (X), and the image data (R', G ', B') is used by a printer (Frontier (trade name) manufactured by FUJIFILM Corporation). The recording of the image on the media and the evaluation of the recorded image by a plurality of evaluators were repeated while changing the values of the parameters of the conversion function f (x). The process flow in the above experiment is summarized as follows.

[0073]

(Equation 10) The conversion function f (x) is obtained by setting the boundary between the predetermined range and the outside of the predetermined range to the parameter u in order to check the appropriate range of the predetermined region.
And is defined by the following equation:

[0074]

[Equation 11] Note that k is a parameter corresponding to the gain under the image data conversion conditions described in the above embodiment, and in this experiment, a density correction coefficient given by the following equation was used (where Σ
G represents the average value of the G density obtained from the image data (R, G, B).

K = 118 / ΣG Further, the parameter k is set to 0.7 and the parameter u is set to “0.
When the values of "7", "0.5" and "0.3" are set, the locus of the conversion function f (x) on the double antilogarithmic chart and the double logarithmic chart is as shown in FIGS. 5 (A) and 5 (B). .

According to the above experiment, the value of the parameter u was substantially
When the value is smaller than 0.3, the deterioration of the image quality of the recorded image is clearly recognized, and when the value (exact value) of the parameter u is approximately 0.3 or more (approximately -0.5 in logarithmic value), the image quality of the recorded image is visually preferable. It was confirmed that. As a result, the inventor of the present application standardized the predetermined area to an area corresponding to a range of approximately 0.3 to 1 (an approximately -0.5 to 0 as a logarithmic value) (that is, an input image data value within a range of 0 to 1). When the image data is converted to a value corresponding to the brightest point in the input image data,
An area corresponding to a range of approximately 0.5 or less in logarithmic value of the input image data value), and the slope of the change of the output image data value with respect to the change of the logarithmic value of the input image data value is approximately 1 only outside the predetermined area. It has been found that if the image data conversion condition is preferably set to 1), the image represented by the image data converted according to the conversion condition has a preferable image quality.

Although an experiment in which only density correction is performed on image data has been described above, with regard to color correction, by defining a predetermined area as described above and setting conversion conditions, the color conversion is performed in accordance with the conversion conditions. It was confirmed by an experiment by the present inventor that the image represented by the converted image data has a preferable image quality.

[0078]

As described above, according to the first and third aspects of the present invention, the brightest point in the input signal is determined based on the correction amount for correcting at least one of the color and the density of the image. The corresponding value is converted to a value corresponding to the brightest point of the output signal, and the slope of the change in the logarithmic value of the output signal value with respect to the change in the logarithmic value of the input signal value is:
The conversion condition is set so as to be approximately 1 except for a predetermined area near the signal value corresponding to the brightest point, and the image data is converted according to the set conversion condition. Color correction and density correction can be performed without causing image quality deterioration such as highlighting in the image, a highlight portion becoming white, and a white background in the image becoming blackish.
It has an excellent effect.

According to a second aspect of the present invention, in the first aspect of the present invention, when the input signal value is normalized within the range of 0 to 1, the input signal value pair is calculated from the signal value corresponding to the brightest point. Since the area corresponding to the range of approximately 0.5 or less in the numerical value is set as the predetermined area, in addition to the above-described effects, the image quality of the image represented by the converted image data can be visually enhanced. Having.

According to a fourth aspect of the present invention, based on a correction amount for correcting at least one of the color and the density of an image,
The value corresponding to the brightest point in the input signal is converted to the value corresponding to the brightest point in the output signal, and the slope of the change in the logarithmic value of the output signal value with respect to the change in the logarithmic value of the input signal value. Includes a first step of setting a conversion condition so as to be substantially 1 except for a predetermined area near a signal value corresponding to the brightest point, and a second step of converting image data according to the set conversion condition. Since the program for causing the computer to execute the processing is recorded on the recording medium, the image quality deteriorates, such as a change in the main gradation of the image, a highlight portion in the image becoming white, and a white background in the image becoming blackish. Color correction and density correction can be performed without the occurrence of the problem.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a schematic configuration of an image processing system according to an embodiment.

FIG. 2 is an image processing system of FIG. 1 in which an information storage medium reading device is used as an input device and a CD-R is used as an output device;
FIG. 4 is a conceptual diagram showing a flow of processing for image data when a writing device is connected.

FIG. 3 is a flowchart showing the contents of a color correction / density correction process for image data.

FIGS. 4A and 4B are diagrams showing an example of a conversion curve on a double logarithmic chart, and FIG.

FIG. 5 shows a conversion function f used in an experiment performed by the present inventors.
(X), (A) is an example of a locus on both antilog charts,
(B) is a diagram each showing an example of a locus on a log-log chart.

[Explanation of symbols]

 Reference Signs List 10 image processing system 14 image data exchange 22 information storage medium reading device 40 CPU 50 hard disk device 52 CD-ROM 54 CD-ROM driver

 ──────────────────────────────────────────────────続 き Continuing on the front page F term (reference) 5B057 AA20 CA01 CA12 CA16 CA18 CB01 CB12 CB16 CB18 CC01 CE11 CE16 CH01 CH11 CH12 DA17 DB06 DC05 5C066 AA01 AA05 BA20 CA17 DD02 EA05 EA07 EA13 EB01 EC02 EC05 EE07 GA04 KE05 KE09 KE17 KF05 5C077 LL01 MP08 PP15 PP16 PP32 PP37 PP46 PP47 PP52 PP53 PQ08 PQ12 PQ18 PQ22 PQ23 RR18 RR21 TT02 TT09 5C079 HB01 HB05 LA12 LA13 LA17 LA23 LB00 MA01 MA04 MA11 NA18 PA03

Claims (4)

[Claims] 1. A value corresponding to a brightest point of an input signal corresponds to a brightest point of an output signal based on a correction amount for correcting at least one of a color and a density of a correction target image. In addition to the conversion into a value, the slope of the change in the logarithmic value of the output signal value with respect to the change in the logarithmic value of the input signal value is substantially 1 except for a predetermined area near the signal value corresponding to the brightest point. An image correction method for setting conditions, and converting image data representing a physical quantity value having a linear relationship with the light intensity of each pixel of the correction target image in accordance with the set conversion conditions. 2. The method according to claim 1, wherein the predetermined area includes an input signal value ranging from 0 to 1.
3. The area according to claim 1, wherein when normalized within the range, the signal value corresponds to a range within approximately 0.5 as a logarithmic value of the input signal value from the signal value corresponding to the brightest point. Image correction method. 3. A value corresponding to the brightest point in the input signal corresponds to the brightest point in the output signal based on a correction amount for correcting at least one of the color and the density of the correction target image. In addition to the conversion into a value, the slope of the change in the logarithmic value of the output signal value with respect to the change in the logarithmic value of the input signal value is substantially 1 except for a predetermined area near the signal value corresponding to the brightest point. Setting means for setting a condition, and conversion means for converting image data representing a value of a physical quantity whose linear relationship with the light intensity of each pixel of the correction target image is linear according to the conversion condition set by the setting means. Image correction device including. 4. A value corresponding to a brightest point in an input signal corresponds to a brightest point in an output signal based on a correction amount for correcting at least one of a color and a density of a correction target image. In addition to the conversion into a value, the slope of the change in the logarithmic value of the output signal value with respect to the change in the logarithmic value of the input signal value is substantially 1 except for a predetermined area near the signal value corresponding to the brightest point. A first step of setting conditions, and a second step of converting image data representing a value of a physical quantity having a linear relationship with the light intensity of each pixel of the correction target image in accordance with the set conversion conditions. A recording medium on which a program to be executed by a computer is recorded.